Synthesis of allylamine plasma polymer films:

correlation between plasma diagnostic and film

characteristics

Laurent Denis,*a Damien Cossement,b Thomas Godfroid,b Fabian Renaux,b Carla

Bittencourt,a Rony Snyders,a,b and Michel Hecqa,b

a
Laboratoire de Chimie Inorganique et Analytique, Université de Mons-Hainaut, Place du

Parc 20, B-7000 Mons, Belgium.

b
Materia Nova Research Center, Parc Initialis, Avenue N. Copernic 1, B-7000 Mons,

Belgium.

*
Corresponding author: Tel: +32(0)65373851, Fax: +32(0)65373841, E-mail: laurent.denis@umh.ac.be

1
Summary

Primary amine-based plasma polymer films (NH2-PPF) were synthesized using plasma

polymerization of allylamine in continuous wave (CW) and pulsed radio-frequency modes.

Plasma chemistry, studied by residual gas analysis mass spectrometry, revealed that the

fragmentation of the chemical precursor is a function of the equivalent power (Peq),

independently of the plasma mode used. X-ray photoelectron spectroscopy combined with

time-of-flight secondary ion mass spectrometry suggest as the precursor fragmentation in the

plasma increases: (i) a decrease of the primary amines concentration in the NH2-PPF (%NH2)

and (ii) an increase of the cross-linking degree. Additionally, for a given Peq, the NH2-PPF

characteristics were found to be independent of the plasma mode used. Therefore, the main

advantage of using pulsed RF processes over CW ones is the possibility to work at very low

Peq what enables low precursor fragmentation, optimization of %NH2 and reduction of the

film cross-linking degree. The chemical composition and the cross-linking degree of the NH2-

PPF synthesized by allylamine plasma polymerization can thus be tailored by adjusting the

equivalent RF power injected in the plasma.

Keywords

plasma polymerization, mass spectrometry, derivatization, ESCA/XPS, secondary-ion mass

spectrometry (SIMS)

2
1. Introduction

Plasma polymerization is an attractive method to synthesize organic thin films referred

to as “plasma polymer films” (PPF). PPF present, among other properties, strong adherence

on numerous surfaces, good mechanical properties and high thermal resistance [1]. Among

the most promising PPF, films containing primary amines (NH2-PPF) have drawn a lot of

interest due to their potential applications ranging from modification of membranes [2-8],

treatment of polymeric beads [9,10] and carbon nanotubes [11] to biomedical applications

[12-18]. Moreover, NH2-PPF have been reported to exhibit relatively good chemical stability

to repeated autoclave cycles and thus resistance to material sterilization, an essential step in

the biomedical research and practice [19]. Several nitrogen-rich organic precursors have been

considered for NH2-PPF synthesis [20-25]. Among them, allylamine (CH2=CH-CH2-NH2) has

extensively been studied since it contains a carbon-carbon double bond susceptible to

polymerize through plasma activation [4,6-10,13-19,22,25-31].

Originally, plasma polymerization was mainly focused on the synthesis of highly

cross-linked films with chemical structure featuring little or no repeating monomer units. In

the last years, however, this technique has shown its strong potential to modify surface

properties of different materials with defined chemical groups. As a consequence, the search

for PPF with controllable and stable chemical composition has triggered the last

developments in the pulsed plasma polymerization area [15]. For instance, power modulation

in pulsed plasma deposition processes allows for a better control over substrate temperature,

organic precursor fragmentation in the plasma and physico-chemical properties of the films

[32,33]. This modulation is defined by the duty cycle (∆) giving the relationship between the

plasma on-time (τon, time during which (re)active species and photons are generated) and the

3
pulse period (sum of τon and the plasma off-time, τoff, time during which the plasma is

switched off).

∆ = τ on (τ on + τ off ) (1)

The equivalent power (Peq) is defined as:

Peq = ∆ ⋅ Ppeak (2)

where, Ppeak is the input power during the plasma on-time. Relatively high peak powers may

frequently be employed but Peq are usually low [33]. Power modulation in pulsed plasma

deposition processes, which is proportional to the duty cycle, has often been used to control

the retention of the organic precursor chemical group in the PPF [19,26,32]. Under such

synthesis conditions, the organic precursor fragmentation increases by increasing Peq.

Nevertheless, even when very low Peq is used, the PPF chemistry is difficult to control

because, as a consequence of the different chemical reactions occurring during plasma

polymerization, a large variety of chemical functions can be incorporated in the PPF. Their

nature and relative concentration depend on the organic precursor as well as on the discharge

parameters. For instance, it has been shown that allylamine PPF can contain different

nitrogen-based chemical functions such as amines, imines and nitriles groups, their relative

concentration depending on the experimental conditions [14,27,28,31].

Among the techniques used for quantifying the chemical composition of solids and

molecules, X-ray Photoelectron Spectroscopy (XPS) is a powerful well-established method

[34]. Nevertheless, in the case of allylamine PPF where different nitrogen-based functions are

4
present, overlapping of contributions in the X-ray photoemission spectrum is likely. For

example, the reported values for the chemical shift of the sp3-hybridized C-N bonds range

from 398 to 398.5 eV; for sp2-hybridized C=N bonds from 400 to 401 eV; and for sp-

hybridized C≡N bonds from 398 to 400 eV [35,36]. Therefore, advanced analysis methods

have to be used for a meaningful identification and quantification of chemical groups. Among

the potential methods, a combination of XPS with chemical derivatization appears to be a

good candidate. It consists in inducing a chemical reaction between a targeted chemical group

of the PPF and a chemical reactant containing at least one atom different from the ones

composing the PPF. Photoelectrons emitted from the element newly introduced can then be

used to quantify the targeted group. For selectively probe –NH2 groups, several reactants such

as 4-trifluoromethyl-benzaldehyde (TFBA) [21,37], pentafluoro-benzaldehyde (PFBA) [23]

and para-chlorobenzaldehyde [38] have been used.

As mentioned before, the synthesis conditions for deposition of PPF with high primary

amines concentration are characterized by low precursor fragmentation. Another consequence

of such experimental conditions is the reduction of the PPF cross-linking degree [39]. And

yet, it has been shown that the control of the cross-linking degree is an important factor for

the optimization of the PPF mechanical and thermal properties [39,40]. Therefore, it is

important to evaluate the cross-linking degree in addition to the plasma and PPF chemistry.

Time-of-Flight Secondary Ion Mass Spectrometry (ToF-SIMS) has been shown to be a

powerful tool to probe structural differences in organic materials [41]. Nevertheless, due to

the ToF-SIMS spectra complexity, statistic methods such as Principal Component Analysis

(PCA) are increasingly used for data analysis [42-48].

5
 In this work aiming to study the synthesis mechanisms of NH2-PPF, the plasma

chemistry is correlated to the film chemical composition and cross-linking degree. Continuous

wave (CW) and pulsed radio-frequency plasma polymerization of allylamine were used to

synthesize the films. The plasma chemistry was studied by residual gas analysis (RGA) mass

spectrometry. The chemical composition of the NH2-PPF was determined by chemical

derivatization XPS using 4-trifluoromethyl-benzaldehyde (TFBA) as labeling molecule while

information about microstructure was deduced from ToF-SIMS data supported by PCA

analysis.

2. Experimental Part

Allylamine (98+ % purity) was purchased from Acros Organics, Belgium. PPF were

deposited on 1cm2-silicon (100) substrates. Prior to PPF syntheses, substrates were

ultrasonically washed in hexane and then rinsed with methanol.

PPF syntheses were carried out in a 350 mm diameter and 450 mm length cylindrical

stainless steel vacuum chamber pumped to a residual pressure of 8.10-7 Torr by a combination

of rotary and turbomolecular pumps connected in series (Fig. 1). The 13.56 MHz exciting

power was applied via a matching network to a water-cooled copper coil, 150 mm internal

diameter and 8 mm tore thickness, located inside the chamber 100 mm in front of the

substrate. A load-lock system allows transferring the substrate inside the reactor from

atmosphere to high vacuum in few minutes. The introduction airlock is separated of the main

chamber by a slide valve. The substrate holder is mounted on a transfer stick and the substrate

kept at floating potential during PPF deposition. A full-range gauge allows controlling the

6
residual pressure before PPF synthesis whereas the process pressure, regulated by a throttle

valve located upstream the turbomolecular pump, is given by a Baratron gauge. The organic

precursor inlet is placed in front of the substrate. Allylamine flow rate was fixed at 10 sccm

during all the experiments and the working pressure regulated at 20 mTorr. Allylamine PPF

were synthesized under pulsed and CW power conditions summarized in Table 1.

A common feature when dealing with plasma polymerization processes is to determine

the PPF deposition regime. For that purpose, the lowest equivalent power condition was

considered (30 W). It has been shown, when the plasma is pulsed, that the deposition regime

depends only on the average energy dissipated per precursor molecule, Emean, in the plasma

[49,50].

Peq Ppeak
E mean = ⋅ τ res = ∆ ⋅ ⋅ τ res (3)
N N

where N is the number of particles resident in the system and τres their resident time, given by

τres = N/Φ, where Φ is the precursor flow rate. Therefore,

Ppeak
E mean = ∆ ⋅ (4)
Φ

It should be emphasized that relation 4 can only be applied if the residence time is much

longer that the pulse period. The calculation of the residence time requires the determination

of N and expression of Φ in number of molecules per second. In our system, the allylamine

flow rate was fixed at 10 sccm corresponding to 4.5x1018 molecules per second. On the other

hand, at a process pressure of 2.67 Pa and for a chamber volume of ~ 4.33x10-2 m3, we obtain

7
N = 2.8x1019 molecules resident every time in the chamber. This gives a residence time of 6.2

s which is much longer than the pulse period of our experiment namely 1.9 msec. An

absorbed power of 1 W/sccm of precursor flow rate corresponds to 15.5 eV of average energy

dissipated per precursor molecule in the plasma [49,50]. For Peq = 30 W, we have Emean = 3

W/sccm (46.5 eV/molecule), this value being consistent with a monomer deficient regime

[51].

Plasma composition was investigated by residual gas analysis (RGA) mass

spectrometry. The mass spectrometer is a quadrupole HAL EQP 500 model supplied by

Hiden Analytical (Warrington, UK). The apparatus is interfaced to the chamber by a 100 µm

extraction orifice located between the coil and the substrate holder. Prior to detection, neutral

species were ionized by electron impact with an electron kinetic energy fixed at 20 eV in

order to limit additional fragmentations in the ion source [52-54]. It should be noted that the

mass spectrometer quadrupole is a mass-dependent system whose transmission function

decreases strongly with the studied mass, m. The manufacturer of the HAL EQP 500 has

found empirically that this mass-dependent instrument function ranges between m-1 and m-2.

In this work, we have used a m-1 mass-dependent function to correct the spectra [55].

XPS experiments were performed on TFBA-derivatized samples. Data were acquired

with a VG-ESCALAB 220iXL spectrometer. A monochromatized Al Kα line (1486.6 eV) was

used as photon source. The overall energy resolution was 0.6 eV. The pressure in the analysis

chamber was 1.5x10-9 Torr. Surface charging effect was compensated using a 6 eV electron

flood gun. Photoemission spectrum background signal was subtracted using the linear

method. Elemental composition was deduced from photoelectron peak areas using respective

photoionisation cross-section calculated by Scofield [56], corrected by the escape depth

8
dependence on the electron kinetic energy (assumed to have the form λ = KE0.6) and by the

analyzer transmission function of the spectrometer.

The derivatization reaction was performed by exposing the PPF to TFBA vapor in a

separated chamber at 3 Torr pressure and room temperature. Scheme 1 shows the reaction

occurring during the –NH2 groups derivatization. The reaction consists in a nucleophilic

addition on carbonyl group (C=O) that converts –NH2 group into imine (Schiff base). Hence,

–NH2 groups are selectively probed by the reagent –CF3 terminal group. Kinetics of this

reaction was investigated as a function of the PPF synthesis conditions. For these

experiments, PPF were TFBA-derivatized using different reaction times. After the

derivatization step, %NH2 was calculated from XPS according to [57]:

[ NH 2 ] ([ F ] / 3)
% NH 2 = = .100 % (5)
[N ] [N ]

where [NH2], [N] and [F] represent respectively the relative concentration of primary amines,

nitrogen and fluorine at the PPF surface. The [NH2]/[N] ratio is referred to as the process

selectivity.

Static ToF-SIMS data of the as-deposited PPF were acquired by a time-of-flight

secondary ion mass spectrometer, using a ToF-SIMS IV instrument from ION-TOF GmbH

(Münster, Germany). An Ar+ 10 keV ion beam at a current of 0.2 pA for samples 1 and 2, and

1 pA for samples 3 and 4, rastered over a scan area of 300 µm × 300 µm was used as analysis

beam. The detection was made in the positive ion mode. For each spectrum the mass scale

was calibrated by using well identified ions namely H+, H2+, H3+, C+, CH+, CH2+, CH3+ and

C2H3+. Five measurements were performed on each sample keeping settings unchanged. The

9
ToF-SIMS spectra were analyzed by the PCA method using the SIMCA-P11 software

supplied by Umetrics, Sweden. The peak intensity of the secondary ions was normalized to

the total ion count before performing the PCA to correct for the differences in total secondary

ion yield from spectrum to spectrum. The normalized spectra were exported from the

acquisition software (IONSPEC v3.16) to the analysis software SIMCA-P11 which mean-

centers and scales the variables. This procedure ensures that the variance in the data are

related to chemical differences between samples and not to artifacts in peak intensity [44,46].

PCA is a multivariate analysis technique aiming at summarizing the variance patterns

within a dataset. The variance in the data describes the differences between samples. In ToF-

SIMS data, these differences come from changes in the relative intensity of the peaks within

samples spectra. Before data processing, results are set in matrix form where the rows are the

samples (i.e. spectra) and the columns are the variables (i.e. peaks). When PCA is applied, a

new set of axes (principal components, PCi) is created defining the directions of the major

variations within the data set. The result of this is a reduction of variables since the original

ones are recombined to define the new PC axis. The large data set is thus reduced to some

variables that are interpreted using specific plots. Two main concepts are then introduced

namely the scores and the loadings. The scores describe the relationship between samples and

account for reproducibility of the surface chemistry results. The loadings define the

contributions of the original variables to the new PCi and describe the variables responsible

for the differences between the samples [44,46].

10
3. Results and Discussion

It has been shown that apart from neutral free radical species, positive ion-molecule

reactions can also have important influence in the PPF deposition, especially if the plasma

polymerization is carried out in CW mode at low input power (~ between 0.5 and 15 W) [58-

61]. Under such plasma conditions, the observation of positively charged oligomers in the

plasma and the absence of such neutral species suggested a cationic chain polymerization

mechanism. During low input power plasma polymerization of acrylic acid, the flux of

positive species arriving at the sample surface revealed the positive ion deposition mechanism

[62]. Similar experiments, considering low input power plasma polymerization of allylamine,

also demonstrated that the ionic mass transported to the surface is sufficient to account for all

the deposit [55]. Nevertheless, for power higher than ~ 15 W, an increased fragmentation of

the organic precursor associated with the loss of positively charged oligomers is observed in

the plasma [63]. Hence, since in this work the power applied in CW or during the plasma on-

time was always ≥ 75 W, it can be assumed that the NH2-PPF growth mechanism occurred

via radicalar species reactions. Consequently, we focused the plasma chemistry analysis on

neutral particles.

The electron impact mass spectrum of allylamine vapor (without plasma) is shown in

Figure 2 (spectrum 2a). Four major peaks are observed at m/z 57, 56, 30 and 28, coming from

α-cleavage reactions of the precursor ([C3H7N]+.). This radical-cation can lose either a

hydrogen radical giving the ([C3H6N]+) signal at m/z 56 or a vinyl radical producing the

([CH4N]+) signal at m/z 30. Subsequent dehydrogenation of the latter yields to ([CH2N]+)

signal at m/z 28. Spectra 2b to 2e show the RGA mass spectra of the plasma as a function of

the experimental conditions. The spectra were recorded from 1 to 60 m/z because no

11
significant signal was detected beyond this value. This result suggests that oligomerization

reactions involving radicalar species (radical-radical or radical-neutral) rarely occur during

the plasma polymerization process. This observation is in good agreement with previous

studies reporting RGA mass spectrometry measurements performed on acrylic acid [51],

propanoic acid [59] and allyl alcohol [60] plasma neutral species.

The different peaks were assigned as follow: m/z 1, 2, 3, ([HX]+) with X varying from 1 to 3;

m/z 15, 16, 17, 18, ([NHX]+) with a contribution from ([CH3]+), ([CH4]+.) and ([H2O]+.) at m/z

15, 16 and 18, respectively; m/z 26, ([CN]+) with a contribution from ([C2H2]+.); m/z 28,

([CH2N]+), ([C2H4]+.) and a possible contribution from ([N2]+.); m/z 30, ([CH4N]+) with a

contribution of ([C2H6]+.); m/z 41 and 42, ([C3H5]+) and ([C3H6]+.), respectively. It is

important to mention that even when low Peq (30 W) is used, complete dehydrogenation of the

chemical precursor is observed at m/z 50. In this region of the spectrum, signals between 58

and 50 m/z correspond to molecular fragments of ([C3HXN]+) as raw chemical formula, with

X varying from 0 to 8. It has been shown that the retention, in the films, of the chemical group

hosted by the precursor behaves in the opposite trend towards the molecule fragmentation in

the gas-phase [51,61]. As a consequence, it can be expected from these mass spectrometry

data that sample 1 contains the highest %NH2 since spectrum 2b, corresponding to the latter

sample, presents the highest signal at m/z = 56. Following the same assumption, sample 4

(spectrum 2e) should present the lowest %NH2 while samples 2 and 3, spectrum 2c and 2d

respectively, should reveal similar chemistry.

In order to probe %NH2, a combination of XPS and TFBA derivatization was

employed. Figure 3 compares the typical XPS survey spectra of as-deposited (spectrum 3a)

and TFBA-derivatized (spectrum 3b) NH2-PPF. Although allylamine is free of oxygen, an

O1s signal at ~ 531.0 eV is observed in both spectra. This is likely the result of post-growth

12
oxidation processes following film exposure to ambient air [64,65]. For the derivatized

sample, appearance of the fluorine (F1s) peak is clearly observed at ~ 685.0 eV. From the

analysis of this latter peak, it is possible to calculate %NH2 as it is mentioned in the

experimental part.

Before the data processing, the quantification method was validated. The first step was

the study of the derivatization reaction kinetics. Figure 4 shows the evolution of %NH2 as a

function of the derivatization duration. The saturation in %NH2 (plateau region) for increasing

derivatization time can be attributed to the complete derivatization reaction of the –NH2

groups. The lowest derivatization time is observed for sample 4 (Peq = 175 W) for which

%NH2 saturates after ~ 90 minutes of derivatization. For samples 2 and 3 (Peq = 75 W),

complete derivatization reaction requires ~ 240 minutes while for sample 1, synthesized at 30

W, the saturation is reached after ~ 960 minutes of derivatization. Hence, in order to be sure

that all –NH2 groups are probed; the latter derivatization time was used for all samples.

In a second step, the depth dependence of the derivatization was probed by modifying

the angle of XPS analysis. We found that, taking into account the experimental uncertainty,

the NH2-PPF composition is independent on the analysis angle used confirming that the XPS

analysis depth, for the excitation energy used, is smaller than the depth of the derivatized

region. As an example, for sample 3, the %NH2 evaluated using the XPS data recorded at a

detection angle (θ) of 90°, 74° and 55° (with reference to the sample surface) were 8.5 ± 1 %,

9.2 ± 1 % and 9.0 ± 1 %.

Table 2 presents the elemental chemical composition of the NH2-PPF, their %NH2 and

the RGA precursor signal (Iallyl) as a function of the synthesis conditions. The results show

13
that %NH2 and Iallyl increase as Peq decreases. Hence, supporting the assumption based on the

RGA data, the sample prepared using the lowest Peq (sample 1) has the highest %NH2 value

while the sample prepared using the highest Peq (sample 4) has the lowest one. The results

show the correlation between the precursor fragmentation and the insertion of primary amines

groups in the PPF. Furthermore, it can be observed that samples 2 and 3 are similar in terms

of chemistry. The latter result is in good agreement with previous studies which have revealed

similar primary amines contents and subsequent DNA immobilization properties for

allylamine PPF synthesized in similar Peq conditions, the power range being lower than the

one studied in this work [13,27].

As already mentioned, in addition to the chemical composition, the PPF cross-linking

degree is also affected by the Peq value. Among the numerous parameters influencing the PPF

microstructure, Emean/M, with M the precursor molar mass, can be correlated to the precursor

fragmentation degree [51,66]. Variation of the power conditions while keeping unchanged the

precursor type and flow rate should thus allow the synthesis of PPF with different cross-

linking degrees. In order to investigate the cross-linked nature of the PPF prepared using

different Peq, the latter were analyzed by ToF-SIMS. Figure 5 shows the PPF positive ToF-

SIMS spectra ranging from m/z = 0 to m/z = 150 as a function of the synthesis conditions.

Except for some differences in relative intensity of the peaks, it is difficult to establish

specific trends.

In order to extract information related to the chemical structure of the NH2-PPF, the

ToF-SIMS data were processed by PCA. Figure 6 shows the positive ToF-SIMS data

presented in a “scores plot” in which each individual point represents a whole ToF-SIMS

spectrum of a given NH2-PPF. The ellipse drawn on the scores plot is the 95% confidence

14
interval calculated from Hotelling's T-square statistic [42]. Two principal components (PCi)

were identified representing together 80% of the data variance: PC1 represents nearly 62% of

the variance while PC2 counts for 18%. The scores plot of PC1 versus PC2, for the PCA

model applied to the ToF-SIMS spectra, distributes the samples according to their

fragmentation pattern. In the scores plot (Fig. 6), the samples 2 and 3 can not be

discriminated, what is in accordance with the XPS results suggesting that these samples are

chemically similar even if they are synthesized in different plasma mode conditions. In

contrast, samples 1 and 4 are clearly discriminated from the other samples.

A scores plot is always associated to a “loadings plot”, the negative scores are

characterized by the negative loadings and the positive scores by the positive loadings. The

loadings reveal the ionized fragments responsible for samples discrimination in the scores

plot. Table 3 shows the PC1 negative and positive loadings whose relative statistical weight

in the model is > 90%. From the scores plot (Fig. 6), sample 1 is clearly different from the

three others having a negative PC1. The characteristic ionized fragments of sample 1,

corresponding to the negative loadings of PC1 (Table 3), are hydrocarbonated ions (CxHy+)

and, in bold type (Table 3), nitrogen-based ions (CxHyN+). On the contrary, for samples 2, 3

and 4, which are characterized by positive values of PC1, the main characteristic ions are

hydrocarbonated (CxHy+), nitrogen-based ions are not statistically important species.

The presence of nitrogen-based fragments for sample 1 indicates a high content of

nitrogen-based chemical groups in the PPF in agreement with the XPS analysis that revealed

the highest %NH2 for this sample synthesized at the lowest Peq (Table 2). Additionally, the

relative chemical composition of the CxHy+ species suggests a lower cross-linking degree for

15
sample 1 compared to the samples 2, 3 and 4 since the ionized fragments detected for this

sample contains an average of 7 carbon atoms compared to 4.5 for samples 2, 3 and 4.

PC1 explains almost 62% of the data variance and clearly discriminates the sample 1

from the others on chemical and structural characteristics. However, distinction between the

samples 2, 3 and 4, characterized by a positive PC1, can also be obtained by considering PC2,

representing 18% of the variance. Similarly than for PC1, Table 4 summarizes the most

important fragments characterizing the positive and negative values of PC2. Sample 4 is

characterized by the negative loadings of PC2 while samples 2 and 3 by the positive ones. The

negative PC2 loadings reveal fragments having in average 3 carbon atoms while, for the

positive loadings, the number of carbon atoms is equal to 12. This result thus suggests a

higher cross-linking degree for sample 4 with regard to samples 2 and 3. Therefore, the latter

have similar structure as well as similar chemical composition. Finally, when comparing the

characteristic ionized fragments of PC2, it can be observed that only the positive PC2 includes

nitrogen-based fragments. This supports the XPS data from which we have deduced a slightly

higher %NH2 for sample 2 and 3 than for sample 4.

4. Conclusions

In this work, CW and pulsed RF plasma polymerization of allylamine were

investigated. The plasma chemistry, studied by RGA mass spectrometry, revealed that the

allylamine fragmentation is strongly dependent on Peq whatever the plasma is sustained in

CW or pulsed mode. In order to evaluate the effect of the precursor fragmentation on the PPF

chemistry, a combination of XPS and TFBA derivatization was used. Additionally, the PPF

16
microstructure was evaluated by identifying the statistically important ionic fragments

composing the ToF-SIMS spectra through PCA analysis.

XPS and ToF-SIMS data revealed, as the precursor fragmentation in the plasma

increases, a decrease of the primary amines concentration in the NH2-PPF (%NH2) while their

cross-linking degree increases, the latter behaving in the opposite trend towards the precursor

mass signal in the plasma. The use of low Peq enables: (i) low precursor fragmentation, (ii)

optimization of the primary amine concentration in the PPF and (iii) reduction of the cross-

linking degree. For a given Peq, the plasma mode (CW or pulsed) has no influence on the PPF

characteristics. Therefore, the main advantage of using pulsed RF processes is the possibility

to work at very low Peq compared to the CW mode.

These findings are of great importance in order to be able to optimize the chemical

selectivity of the primary amine-based PPF as well as their cross-linking degree.

Acknowledgements

This work is supported by the « Fonds pour la Formation à la Recherche dans

l’Industrie et dans l’Agriculture » (F.R.I.A.) and the Belgian Government through the « Pôle

d’Attraction Interuniversitaire» (PAI, P6/08, “Plasma-Surface Interaction”, Ψ).

17
References

[1] “Plasma Technology, vol. 3 – Plasma Polymerization Processes”, H. Biederman, Y.

Osada, Eds., Elsevier Science Publishers, Amsterdam 1992.

[2] H. Matsuyama, K. Hirai, M. Teramoto, J. Membr. Sci. 1994, 92, 257.

[3] R. C. Ruaan, T. H. Wu, S. H. Chen, J. Y. Lai, J. Membr. Sci. 1998, 138, 213.

[4] H. I. Kim, S. S. Kim, J. Membr. Sci. 2001, 190, 21.

[5] G. Pozniak, I. Gancarz, M. Bryjak, W. Tylus, Desalination 2002, 146, 293.

[6] I. Gancarz, G. Pozniak, M. Bryjak, W. Tylus, Eur. Polym. J. 2002, 38, 1937.

[7] P. Hamerli, Th. Weigel, Th. Groth, D. Paul, Biomaterials 2003, 24, 3989.

[8] P. Hamerli, Th. Weigel, Th. Groth, D. Paul, Surf. Coat. Technol. 2003, 174-175, 574.

[9] G. Oye, V. Roucoules, A. M. Cameron, L. J. Oates, N. R. Cameron, P. G. Steel, J. P. S.

Badyal, B. G. Davis, D. Coe, R. Cox, Langmuir 2002, 18, 8996.

[10] G. Oye, V. Roucoules, L. J. Oates, A. M. Cameron, N. R. Cameron, P. G. Steel, J. P. S.

Badyal, B. G. Davis, D. M. Coe, R. A. Cox, J. Phys. Chem. B 2003, 107, 3496.

[11] Q. Chen, L. Dai, M. Gao, S. Huang, A. Mau, J. Phys. Chem. B 2001, 105, 618.

[12] L. Dai, H. A. W. StJohn, J. Bi, P. Zientek, R. C. Chatelier, H. J. Griesser, Surf. Interface

Anal. 2000, 29, 46.

[13] Z. Zhang, Q. Chen, W. Knoll, R. Foerch, R. Holcomb, D. Roitman, Macromolecules

2003, 36, 7689.

[14] Q. Chen, R. Förch, W. Knoll, Chem. Mat. 2004, 16, 614.

18
[15] R. Förch, Z. Zhang, W. Knoll, Plasma Process. Polym. 2005, 2, 351.

[16] Z. Zhang, W. Knoll, R. Förch, Surf. Coat. Technol. 2005, 200, 993.

[17] T. B. Ren, Th. Weigel, Th. Groth, A. Lendlein, J. Biomed. Mater. Res. Part A 2008, 86A,
209.

[18] A. L. Hook, H. Thiessen, J. Quinton, N. Voelcker, Surf. Sci. 2008, 602, 1883.

[19] A. Harsch, J. Calderon, R. B. Timmons, G. W. Gross, J. Neurosci. Methods 2000, 98,

135.

[20] T. R. Gengenbach, H. J. Griesser, J. Polym. Sci. Pol. Chem. 1999, 37, 2191.

[21] A. Choukourov, H. Biederman, D. Slavinska, M. Trchova, A. Hollander, Surf. Coat.

Technol. 2003, 174-175, 863.

[22] A. G. Shard, J. D. Whittle, A. J. Beck, P. N. Brookes, N. A. Bullett, R. A. Talib, A.

Mistry, D. Barton, S. L. McArthur, J. Phys. Chem. B 2004, 108, 12472.

[23] J. Kim, D. Jung, Y. Park, Y. Kim, D. W. Moon, T. G. Lee, Appl. Surf. Sci. 2007, 253,
4112.

[24] Y. Martin, D. Boutin, P. Vermette, Thin Solid Films 2007, 515, 6844.

[25] G. S. Malkov, I. T. Martin, W. B. Schwisow, J. P. Chandler, B. T. Wickes, L. J. Gamble,

D. G. Castner, E. R. Fisher, Plasma Process. Polym. 2008, 5, 129.

[26] H. Schönherr, M. T. van Os, R. Förch, R. B. Timmons, W. Knoll, G. J. Vancso, Chem.

Mat. 2000, 12, 3689.

[27] A. Choukourov, H. Biederman, D. Slavinska, L. Hanley, A. Grinevich, H. Boldyryeva,

A. Mackova, J. Phys. Chem. B 2005, 109, 23086.

19
[28] M. Tatoulian, F. Bretagnol, F. Arefi-Khonsari, J. Amouroux, O. Bouloussa, F. Rondelez,
A. J. Paul, R. Mitchell, Plasma Process. Polym. 2005, 2, 38.

[29] U. Oran, S. Swaraj, A. Lippitz, W. E. S. Unger, Plasma Process. Polym. 2006, 3, 288.

[30] S. Igarashi, A. N. Itakura, M. Toda, M. Kitajima, L. Chu, A. N. Chifen, R. Förch, R.

Berger, Sens. Actuator B - Chem. 2006, 117, 43.

[31] M. Lejeune, F. Bretagnol, G. Ceccone, P. Colpo, F. Rossi, Surf. Coat. Technol. 2006,
200, 5902.

[32] C. R. Savage, R. B. Timmons, J. W. Lin, Chem. Mat. 1991, 3, 575.

[33] R. B. Timmons, A. J. Griggs, “Pulsed Plasma Polymerizations”, in: Plasma Polymer

Films, H. Biederman, Ed., Imperial College Press, London 2004.

[34] “Photoelectron Spectroscopy: Principles and Applications”, 3rd edition, S. Hüfner, Ed.,
Springer, Berlin 2003.

[35] A. K. M. S. Chowdhury, D. C. Cameron, M. S. J. Hashmi, Surf. Coat. Technol. 1999,

112, 133.

[36] D. Q. Yang, E. Sacher, Surf. Sci. 2003, 531, 185.

[37] I. Losito, E. De Giglio, N. Cioffi, C. Malitesta, J. Mater. Chem. 2001, 11, 1812.

[38] R. Snyders, O. Zabeida, C. Roberges, K. I. Shingel, M. P. Faure, L. Martinu, J. E.

Klemberg-Sapieha, Surf. Sci. 2007, 601, 112.

[39] R. Prikryl, V. Cech, L. Zajickova, J. Vanek, S. Behzadi, F. R. Jones, Surf. Coat. Technol.
2005, 200, 468.

[40] D. D. Burkey, K. K. Gleason, Chem. Vapor Depos. 2003, 9, 65.

20
[41] L. Lianos, C. Quet, T. M. Duc, Surf. Interface Anal. 1994, 21, 14.

[42] M. S. Wagner, D. G. Castner, Langmuir 2001, 17, 4649.

[43] M. von Gradowski, M. Wahl, R. Förch, H. Hilgers, Surf. Interface Anal. 2004, 36, 1114.

[44] M. S. Wagner, D. J. Graham, B. D. Ratner, D. G. Castner, Surf. Sci. 2004, 570, 78.

[45] M. von Gradowski, B. Jacoby, H. Hilgers, J. Barz, M. Wahl, M. Kopnarski, Surf. Coat.
Technol. 2005, 200, 334.

[46] D. J. Graham, M. S. Wagner, D. G. Castner, Appl. Surf. Sci. 2006, 252, 6860.

[47] B. J. Tyler, Appl. Surf. Sci. 2006, 252, 6875.

[48] B. J. Tyler, G. Rayal, D. G. Castner, Biomaterials 2007, 28, 2412.

[49] M. Bauer, T. Schwarz-Selinger, H. Kang, A. von Keudell, Plasma Sources Sci. Technol.
2005, 14, 543.

[50] M. Bauer, T. Schwarz-Selinger, W. Jacob, A. von Keudell, J. Appl. Phys. 2005, 98,
073302.

[51] S. A. Voronin, M. Zelzer, C. Fotea, M. R. Alexander, J. W. Bradley, J. Phys. Chem. B

2007, 111, 3419.

[52] “Cold Plasma in Materials Fabrication – From Fundamentals to Applications”, A. Grill,

Ed., IEEE Press, New York 1994.

[53] M. J. Vasile, G. Smolinsky, Int. J. Mass Spectrom. 1976, 21, 263.

[54] R. Manory, A. Grill, U. Carmi, R. Avni, Plasma Chem. Plasma Process. 1983, 3, 235.

21
[55] A. J. Beck, S. Candan, R. D. Short, A. Goodyear, N. St J. Braithwaite, J. Phys. Chem. B
2001, 105, 5730.

[56] J. H. Scofield, J. Electron Spectrosc. Relat. Phenom. 1976, 8, 129.

[57] F. Truica-Marasescu, M. R. Wertheimer, Plasma Process. Polym. 2008, 5, 44.

[58] D. B. Haddow, R. M. France, R. D. Short, J. W. Bradley, D. Barton, Langmuir 2000, 16,

5654.

[59] L. O’Toole, A. J. Beck, A. P. Ameen, F. R. Jones, R. D. Short, J. Chem. Soc. Faraday

Trans. 1995, 91, 3907.

[60] L. O’Toole, C. A. Mayhew, R. D. Short, J. Chem. Soc. Faraday Trans. 1997, 93, 1961.

[61] L. O’Toole, R. D. Short, J. Chem. Soc. Faraday Trans. 1997, 93, 1141.

[62] S. Candan, A. J. Beck, L. O’Toole, R. D. Short, A. Goodyear, N. St J. Braithwaite, Phys.

Chem. Chem. Phys. 1999, 1, 3117.

[63] M. R. Alexander, F. R. Jones, R. D. Short, J. Phys. Chem. B 1997, 101, 3614.

[64] T. R. Gengenbach, Z. R. Vasic, R. C. Chatelier, H. J. Griesser, J. Polym. Sci. Pol. Chem.

1994, 32, 1399.

[65] T. R. Gengenbach, Z. R. Vasic, S. Li, R. C. Chatelier, H. J. Griesser, Plasmas Polym.

1997, 2, 91.

[66] “Plasma Polymerization”, H. Yasuda, Ed., Academic Press, Orlando 1985.

22
List of Figures

Figure 1

23
Figure 2

24
Figure 3

25
Figure 4

26
Figure 5

27
Figure 6

28
Scheme

Scheme 1

29
List of Tables

Sample Equivalent
Power conditions Plasma mode
number power conditions
1 150 W (peak power) – 20 % (duty cycle) Pulsed 30 W
2 225 W (peak power) – 33 % (duty cycle) Pulsed 75 W
3 75 W CW 75 W
4 175 W CW 175 W

Table 1

30
 Elemental composition of the
Equivalent power Iallyl
derivatized-PPF %NH2 (%)
conditions (arbit.units)
[C] (%) [N] (%) [O] (%) [F] (%)
30 W – Pulsed 72.8 14.2 4.3 8.7 20,4 270
75 W – Pulsed 78.3 15.1 2.7 3.9 8,6 30
75 W – CW 78.4 15.2 2.5 3.9 8,5 35
175 W – CW 80.1 14.6 2.4 2.9 6,6 5

Table 2

31
 Negative loadings of PC1 Positive loadings of PC1
C7H12+ C9H13N+ CH4N+ C3H3+ C2H3+ C5H4+
C7H11+ C8H12+ C3H7N+ C4H3+ C5H5+ CH3+
C6H10+ C6H11+ C4H8+ C6H5+ C5H2+ C11H9+
C5H10N+ C9H13+ C9H12+ C3H2+ C10H8+ C6H3+
C5H10+ C9H14+ C2H6N+ C5H3+ C4H5+ CHO+
C8H13+ C7H10+ C8H11+ C3H5+ C4H2+ C3H4+
C3H8N+ C4H6N+ NH4+ C2H2+ C9H7+ C2H+

Table 3

32
 Negative loadings of PC2 Positive loadings of PC2
C10H13+ C2H3O+ C5H11+ C15H14+ C14H12+ C11H8+
C5H9+ C4H7+ CHO+ C8H8+ C10H10+ C13H11+
C3H5O+ C2H4+ CH3+ C13H12+ C14H13+ C16H8+
C+ CH3O+ C3H6+ C15H13+ C13H13+ C15H10+
C6H13+ CH2+ C3H4+ C16H12+ C9H10+ C15H12+
CH+ C3H7+ C4H9+ C4N+ C4HN+ C12H13+

Table 4

33
Figure captions

Figure 1. Experimental Set-up - 1: introduction airlock; 1’: substrate holder; 2: slide valves;

3: water-cooled radiofrequency copper coil; 4: optical spectrometer; 4’: optical fibre; 4”:

collimator network; 5: mass spectrometer; 6: organic precursor inlet; 7, 7’: turbomolecular

pump and throttle valve; 8: rotary pump; 9: Baratron gauge; 10: tap valves.

Figure 2. (a) Electron impact mass spectrum of allylamine vapor, (b-e) RGA mass spectra of

allylamine plasma varying the Peq conditions.

Figure 3. Typical XPS survey spectra of allylamine plasma polymer films (a) as-deposited

and (b) TFBA-derivatized.

Figure 4. %NH2 as a function of the derivatization duration (the dotted lines are added to

guide the eye in underlining the time at which the reaction is complete).

Figure 5. Positive ToF-SIMS spectra of the allylamine PPF varying the Peq conditions.

Figure 6. “Scores plot” of the PPF showing samples separation according to PCA analysis of

the ToF-SIMS data.

34
Scheme title

Scheme 1. Chemical gas-phase reaction occurring during TFBA derivatization of the PPF –

NH2 groups.

Table titles

Table 1. Experimental conditions under which allylamine PPF were synthesized.

Table 2. Elemental chemical composition and %NH2 of the PPF with regard to the precursor

RGA signal in the plasma (Iallyl) varying the Peq conditions.

Table 2. PC1 negative and positive loadings. The characteristic ionized fragments are

hydrocarbonated (CxHy+), nitrogen-based (CxHyN+) and, at a lesser extent, oxygen-based

(CxHyO+). The latter are likely the result of post-growth oxidation processes following

samples exposure to atmosphere.

Table 3. PC2 negative and positive loadings. The ionized fragments are hydrocarbonated

(CxHy+), nitrogen- (CxHyN+) and oxygen-based (CxHyO+).

35
Text for the “Table of Contents”

Plasma-polymerized surfaces presenting high primary amines concentration are of

importance for many biomedical applications. We have studied the plasma polymerization of

allylamine by correlating mass spectrometry diagnostic with the chemical composition

(XPS/TFBA derivatization) and cross-linking degree (ToF-SIMS/PCA) of the films. We have

demonstrated that these features depend directly on the equivalent RF power injected in the

plasma but are not affected by the plasma mode used (pulsed or continuous wave).

Graphic for the “Table of Contents”

36